Since the end of the 1960s, price limits has been employed by many futures markets and stock markets around the world. In the literature, the effectiveness of price limits is still under debate. The main purpose of price limit is to reduce price volatility. The rationale for supporting price limits is referred to overreaction hypothesis. Proponents asserted that traders are prone to overreact to new information. Therefore, asset prices may encounter large changes so that they deviate from fundamental values. In this situation, price limits provide a cooling-off period for traders to reassess the intrinsic asset value. Volatility is then reduced. Opponents for price limits argued that price limits may have negative effects on the financial markets. These are known as the delayed price discovery hypothesis (or information hypothesis), volatility spillover hypothesis, and trading inference hypothesis. Information hypothesis claimed that the main effect of price limits is just to delay the process of price discovery. If limit moves are present and intrinsic value of asset price falls outside the current price variation range, all trading will suspend and will resume when the intrinsic value lies within the new price variation range. The role of price limits just delay the trading activity. Price continuations occur in the next trading day following limit moves. As a result, this will enhance the autocorrelation of asset returns. Volatility spillover hypothesis states that volatility is increased in the subsequently trading days after limit moves due to imbalance orders caused by price limits. As to the trading interference, it claims that trading activity will be interfered once asset prices hit the limits. Therefore, the trading activity in the subsequent trading days will increase. In this paper, the time series properties and several hypotheses described above are examined in the framework of agent-based artificial stock market in which trader' behavior is modeled by genetic programming
